This work has received financial support from the LabEx DynamiTe (ANR-11-LABX-0046), as part of the “Investissementsd’Avenir” program.

2Macro-regional integration therefore remains an important and legitimate object of study. First of all, many researchers converge towards the conclusion that ultimately, globalization has neither abolished distance, nor detracted from the positive effects of geographic proximity. Spatial constraints continue to affect trade as well as the density and intensity of interactions between societies. Even in a globalized world, pervaded with networks that disrupt traditional territorialities, building relations with one’s neighbors can be useful, if not inevitable. In addition, the crises undergone by the E.U. must not detract us from what is currently at play in other parts of the world. For example, since the 2000s, South-East Asian countries have been working on the construction of a consolidated regional mechanism, following a different model than Europe’s. In this context, what do we mean by macro-regional integration? Is it only relevant to governments engaged in bi- or multilateral treaties? What is at play on other scales, in particular between non-institutional players? Does the ending of an international agreement cancel everything that has been built at the grassroots by other players such as civil society, local authorities, businesses, not-for-profit operators, etc.?

3Multi-state macro-regional constructs deserve to be explored by geographers. This theme has long remained under-studied by this discipline, while producing a remarkable wealth of theoretical and empirical studies in other fields. As well as being underrepresented, studies on this issue remain scattered as there is currently no research team dedicated to this theme. In contrast, in economics for example, some research teams have specific focal areas built on the theme of macro-regional integration: examples include the “Fondation pour les Études et Recherches sur le Développement International” (Foundation for Studies and Research on International Development, or DERDI) in Clermont-Ferrand (France), and the “Groupe de Recherches sur l’Intégration Continentale” (Research Group on Continental Integration, or GRIC) in Montréal (Canada).

4In addition, geography suffers from a certain theoretical weakness: attempts at generalizing and organizing concepts have been more precocious and much more advanced in other disciplines (Mareï, Richard, 2018). And yet, because of their undisputed expertise on themes such as regions and spatial integration, geographers could have a lot to say on macro-regional integration.

The importance of regional integration in other disciplines

5This theme is mainly studied by three disciplines: economics, political science and international law. While their input is notable, it also appears quite scattered. There is no consensus among researchers as to the theoretical frameworks that should be used, or as to the definition of certain terms. Words like “regionalization”, “regionalism” and “regional integration” are omnipresent but their meaning often remains fuzzy. Finally, some notions are used without any prior definition, which causes the question under consideration to become unclear.

6Some pioneering disciplines have brought a major contribution to the knowledge of macro-regional integration. However, irrespective of their scientific quality, existing studies present certain biases. Most studies in political science, international relations and international law follow a state-centric approach, presenting governments as the main actors (Hoffman, 1990; Hoffmann and Keohane, 1991) and states as the main links, if not sole ones, in regionalism and integration. An example can be found in the definition of regional integration written by a political scientist in the Oxford Research Encyclopedia:

“Regional integration theory seeks to explain the establishment and development of regional international organizations. Key questions are why and under which conditions states decide to transfer political authority to regional organizations; how regional organizations expand their tasks, competencies, and members; and what impact they have on states and societies in their regions. Whereas regional integration theory started with a broad comparative regional and organizational scope in the 1950s and 1960s, it has since focused on European integration and the European Union” (Oxford Research Encyclopedias, Frank Schimmelfennig, International Political Economy, Political Economy, 2018).

7In international law, integration occurs when states jointly create common institutions under the framework of a regional agreement. Consequently, NAFTA is not an example of macro-regional integration because its member countries have not created any institutions. In economics, Bela Balassa’s work serves as a benchmark for measuring levels of integration on a scale ranging from free trade agreements to customs unions, single market, etc. (Balassa, 1961; Baldwin 1993; Baldwin, 1997; Frankel, et al., 1995; Krugman, 1991 and 1993). Such approaches tend to place the emphasis on institution-led initiatives and dynamics. For instance, a regional trade agreement signed by governments is regarded as sufficient to create a region. This state-centric approach is dismissed, among others, by the sociology of international relations, which focuses more on other actors of varying levels and natures (Saurugger, 2010). The neo-functionalist approach to integration avoids the pitfall of state-centrism by observing both state and non-state actors, as well as supranational institutions (Haas, 1958). Along with other approaches, proponents of “new regionalism” within the broader movement of social constructivism present studies in which states, economic actors (markets) and civil societies are all equally taken into account, in order to understand and analyze integration processes (Hettne, Söderbaum, 2000; Slocum, Van Langenhove, 2004).

8Geographical space is generally absent from the studies produced in these disciplines. It is regarded at most as a distant background with none or little influence (Mansfield, Milner, 1999). Few politicians have placed space at the center of their work. Those who have, including Bruce Russett (1967) and William Thompson (1973), only had a sparse following. While these researchers considered the notion of “international region” in the context of their reflections on international regimes, the concept of region remains unclear in their work because the principle of spatial contiguity, considered a necessary condition by geographers to refer to a region, is for them only one possible but not necessary criterion among many others. Robert Keohane and Joseph Nye (2000) have long raised the question of the distance required to use the term “region”. According to these authors, the notion of distance can be subjected to variations as it can embrace diverse scales from the regional to the global. However, they do not specify the threshold beyond which a distance can no longer be regarded as regional.

9Some weaknesses can be pinpointed in these approaches. For example, internationalists often postulate the existence of a given region without taking the trouble to demonstrate it (Vayrÿnen, 2003), or to verify that it exists on a functional level. They believe that a trade agreement suffices to establish a region – as shown for instance by the claims of the World Bank and World Trade Organization (WTO). For the former, "a regional trade agreement (RTA) is a treaty between two or more governments that define the rules of trade for all signatories". For the latter, "regional trade agreements [...] are reciprocal preferential trade agreements between two or more partners". The idea of spatial continuity, which is inherent in the geographers’ understanding of the region, is absent from these definitions. Moreover, some publications in economics rely on presupposed divisions of the world without any critical distance, considering so-called “regional” units on the same level although they are ultimately not very comparable. This has for instance been the case for years in the WTO’s annual reports on international trade (Siroën, 2000).

10Conventional analyses are often mono-sectoral in nature because they are limited to one field, which results in an overly simplistic representation of reality. In economics, few authors take a global view of macro-regional integration (GEMDEV, 1999). Some economists follow other paths. For instance, Henri Regnault (2008) extends his analyses beyond the strictly economic sphere of trade, questioning the relevance of regionalism’s traditional classifications. In his work, a “regional” trade system derives from what he calls “a desire to live together” or, at the very least, from the awareness of a “duty to live together” that makes this system a global geopolitical project. Another example can be found in the work of P. Hugon, who stresses that macro-regional integration can have a spontaneous, non-institutional character. Hugon also mentions that this process is not limited to the economy and trade, and that it can go along with joint initiatives on security and development, for example. Finally, this author shows that macro-regional integration can be the de facto product of “practices led by actors who form commercial, financial, cultural and technological networks in regional territories”, and who seek, among other things, to take advantage of complementarities between neighboring territories (Baldwin, Venables, 1995; Baldwin et al., 1997; Hugon, 1999 and 2003).

Macro-regions in geography: a vibrant area of research

11In geography, large-scale regional units are a relatively recent object of study – although geographers have for many years contributed to the spatial definition of continents, with some pioneering works in this field (Claval, 1968; Didelon, 2013). The variety of regional perspectives on the world in recent geographical studies testifies to the vitality of the discipline. However, it also reflects a lack of consensus on concepts and methods (Mareï, Richard, 2018). In essence, the state of current research can be summarized into three main types of geographical approaches to so-called “global” regions, which will be examined in this section. We will leave out the studies that seek to identify large homogeneous regions, because the usefulness and relevance of the very notion of homogeneity in regional geography has long been questioned (Juillard, 1961). While this notion was already difficult to establish on an intra-state level, it proves even more challenging when considering multi-state regions (Claval, 1968; Didelon, 2013).

12The first approach regionalizes the world based on continental divisions. It consists in delineating large spatial units a priori on the world map, and then studying their internal organization, without justifying their boundaries or focusing on macro-regional integration. This bias is very common, and it provides the foundations of a significant part of the literature in so-called “regional” geography. This approach was chosen by many of the authors who undertook to study large continental units. This is the case in many geographies of Europe (Gottmann, 1962; Frémont, 1996; Vandermotten, Dézert, 2008), which start from the more or less explicit hypothesis that there is a region called “Europe”. This region is postulated as an already-there. This way of proceeding runs the risk of excessively naturalizing the geographical units thus defined, even though the authors often stress that the boundaries of Europe are a pure matter of representation and convention, and that they are ultimately impossible to define. Other parts of the world can also be postulated a priori as regions, as shown by the many geographies of Africa, Asia, etc. In most of these publications, the chosen region – which generally derives from tradition and from the persistence of a traditional continental division of the world – is postulated more than demonstrated. In a slightly different and not so distant spirit, one can mention all the universal geographies (written and/or edited by Malte-Brun, Reclus, Vidal and Gallois, Quillet, Brunet…) published since the nineteenth century, whose authors also imposed world divisions with little or no justification, based on ad hoc criteria (Arrault, 2007; Didelon 2013).

13The second approach also consists in dividing the world a priori into regional multi-state groupings, without any particular justification of their boundaries, but with the explicit aim of studying the tools and effects of inter-state cooperation and potentially integration. This is for instance what Christian Taillard (2004) does in his seminal collective book on East Asia. This cross-disciplinary publication addresses the issue of regionalization from a global and multi-sector perspective (economy, tourism, geopolitics, economic and political regionalism, Chinese diaspora, labor migration, transnational cooperation...). The authors point out the methodological problems and the recurring imprecisions of the vocabulary associated with the themes of regionalization and macro-regional integration. However, ultimately, they hardly ever question the relevance of “East Asia” as a geographical category. This regional unit is postulated and based on traditional boundaries, since the authors choose to include all the countries located on the western side of the Pacific, including Southeast Asia, Korea, China and Japan.

14Geographers like François Taglioni and Jean-Marie Théodat (2007) also choose to divide macro-regions a priori in their book on Latin America, where they study a number of internal mechanisms that they associate with regionalism and macro-regional integration. In the same vein, with sometimes geopolitical approaches, we can mention the work of A. Rückert (2018), C. Girault (2009) on South America, L. Medina on Central America (Medina-Nicolas, 2007 and 2009), V. Bachmann and J. Sidaway on Africa (Bachman and Sidaway, 2010; Bachmann, 2011; Sidaway, 1998), F. Mohamed-Chérif and C. Ducruet on North Africa (2016) or J. Poon on East Asia (2001). Some of these regional geographies follow a regional approach (cutting across different sectors), while others adopt a sectoral perspective (Carrizo, Velut, 2018; Palle, 2018). There is also a very abundant literature on the dimensions and geographical effects of European construction on different scales, which tends to focus on recurring themes such as cross-border cooperation (Reitel, Wassenberg, 2015), the tools and impacts of the E.U.’s regional policy (Elissalde, Santamaria, 2008 ; Faludi, 2008; Boulineau, 2017) and those of its cohesion policy (Baudelle, Mérenne-Schoumaker, 2011; Baudelle, le Bihan, 2017), territoriality and governance (Mamadouh, Van Der Wüsten, 2009), etc.

15The third approach does not bring forward an a priori definition of large geographical units. Starting for example from the hypothesis that geographical proximity is a determining factor in the geographical distribution of social practices, geographers focus on the distribution of certain indicators and/or on the spatial distribution of flows and exchanges of all kinds (trade, migration, tourism, investment and other financial flows, etc.). In the first case, they seek to identify large homogeneous global regions, sometimes using fine quantitative methods (Beckouche, Grasland et al., 2008). In the second case, they measure the intensity of relations and interactions between spatial units – often using modelling – or reconstruct the geography of value chains. Finally, they devise maps documenting all observed and measured facts. When statistical corpuses are insufficient, geographers resort to more qualitative methods such as multi-site observation (Retaillé, 2012) or field data collection (Marfaing and Wippel, 2004). The resulting cartographic outputs then enable scholars to reveal, where relevant, a regional concentration of certain social facts or a potential interdependency between territories, and to draw conclusions as to the potential existence of large-scale spatial systems (Reynaud, 1981; Dollfus, 1994; Poon, 1997; Taillard, 2004; Didelon and Grasland, 2007; Beckouche, 2008; Richard, Zanin, 2009; Grasland, Van Hamme, 2010). Such inductive approaches have the advantage of revealing potential regional aggregates, while never presuming their existence. In most cases, the boundaries of these units appear disconnected from both traditional divisions of the world such as continents, and regional trade agreements. These approaches are of course reminiscent of Fernand Braudel’s notion of "world-economies", which he described as examples of spatial integration on a macro-regional scale (Braudel, 1986). There is however one difference: “world-economies” are not all “regional” in the literal sense of the term, since Braudel never presented spatial continuity as an explicit and absolute criterion.

Returning to the geographical fundamentals

16In addition to empirical approaches, the notion of macro-regional integration needs to be clarified, generalized and above all, conceptualized. This effort can build on two fundamental geographical notions: spatial integration and regions. Working from these notions, we can then bring forward definitions of regionalization, regionalism and macro-regional integration, based on the premise that these notions are not interchangeable.

Region and spatial integration: two fundamental geographical notions

17All geographers are familiar with the notion of region. Regionalizing has long been one of the fundamental tasks of geography (Claval, 1995), although the region no longer occupies as central a position in the discipline as it once did. The notion can be defined in several ways: a more or less arbitrary spatial division of a larger whole; a political and administrative management unit (or even an operational planning area) (Montello, 2003); or a functional reality (Juillard, 1962; Seys, 2019). According to the latter definition, it is a continuous portion of space, without presumption of size, defined by its internal homogeneity (Haggett, 1973) and/or where internal interactions (exchanges, movements, social relations, etc.) are stronger than interactions with the external environment (Hartshorne, 1939; Juillard, 1962; Boudeville, 1964; Montello, 2003). The region is described as a coherent, highly integrated, continuous space, usually identified on an infra-state level. Regardless of the chosen definition, the region as seen by geographers is above all a social construct, like any form of spatial organization (Allen et alii, 1998; Paasi, 2010). As such, it is dynamic: its boundaries are shifting or even blurred (Ruffray, 2007). It can be done and undone, but it forms an autonomous and coherent whole with a degree of stability over time (Claval, 1968).

18At the highest level of generality, spatial integration can refer to a growing interdependence between the constituent parts of a whole, or to the inclusion of a spatial unit into an already existing system, causing a reciprocal transformation of the primordial elements (Nonn, Martin, 1980). The result of spatial integration is the construction of a new geographical reality (Marcuse, 1997). This new reality is more than the sum of its parts and is free from internal barrier effects, but it does not necessarily form a homogeneous whole (Lee, 2009). This definition builds on other related notions such as accessibility and connectivity. It has the advantage of functioning on various scales: it can refer to interactions between countries (Dabinett, Richardson 2005), intra-state territorial units (Anderson, Wever 2003; Decoville et al., 2013), cities (van Oort, Burger, Raspe 2010), intra-urban spaces (Hansen, Serin 2007; Sohn, Reitel, Walther 2009), etc. Spatial integration can be sectoral (i.e. restricted to one sector), multi-sectoral or even global (affecting economic, political and social interactions in the broadest sense). Some authors also use the notion of spatial integration to refer to the reduction of differences between territories (De Boe, Grasland, Healy, 1999): this type of integration is synonymous with convergence and homogenization. Both types of integration (interaction or convergence) are not necessarily correlated (Topaloglou et al., 2005). When integration derives from the growth of internal interactions, the resulting integrated spatial system is made up of complementary units that form a heterogeneous whole. Such a system is often asymmetrical and it is structured by center-periphery relations (Reynaud, 1981).

Regionalism and regionalization according to geographers

19On a macro level, regional integration (see Part IV) should be distinguished from regionalism and regionalization (see Figure 1). The definition of regionalism poses relatively few problems. “Regionalism” refers to situations where one or more actors decide to prioritize relations with their vicinity in their external actions. Regionalism is not a spontaneous process. On the contrary, it involves conscious and planned action, a strategy implemented at a given moment by actors located in the same part of the world. One can use the notion of regionalism, for example, when several neighboring states decide to remove the border-related barriers that separate them. Regionalism is often led by states that wish to give their local relations a more organized or better regulated character and to develop their cooperation in certain areas. Strictly speaking, regionalism translates into the signing of agreements that provide the cooperation with a structured framework, defining its limits and objectives.

20When applied to spatial dynamics, regionalization refers to the regional concentration of exchanges, i.e. when exchanges between spatial units located in the same part of the world develop faster than exchanges with the rest of the world. These territorial units may be countries or sub-state entities. All flows or other types of relations can become regionalized: tourist flows, migration or international mobility, development aid, communications, investment, etc. For example, migratory flows develop within ensembles of neighboring countries, thus forming regional units. In the Americas, migration from South to North America has contributed to the emergence of migratory connections across the border between Mexico and the United States. In the Middle East, mobility between the Arab Gulf States and the Mashreq countries has led to the emergence of a regional migration system (Vignal, 2017). Similarly, the overwhelming majority of migrants in Europe come from neighboring countries. Although some distances have been abolished by new ways of accessing information and circulating, and although globalization has also generated new long-distance migratory systems, geographical proximity remains a determining factor in the formation of regional migratory connections or systems (Przybyl, 2018).

21The regionalization of the world is a well-documented phenomenon in certain fields such as the international goods trade, thanks to these sectors’ more effective statistical tools. Uneven degrees of regionalization can be observed. According to the UN, in 2018, an average of 68% of European countries’ foreign trade was with other European countries (UN, 2019). In other parts of the world, the “regional” concentration of international trade is significantly less pronounced: about 37% in East Asia, 18% in Sub-Saharan Africa, 15% in Latin America and the Caribbean, etc. Trade statistics also show that regionalism and regionalization should not be confused, as there is no systematic correlation between the two phenomena. International trade is often not very intense between neighboring countries, even when they belong to the same trade agreement. From this perspective, European Union with its 28 members is an exception: over 61% of member countries’ international trade is completed within the Community, in comparison for example with 40% of internal trade in NAFTA, 22% in ASEAN and 16% in Mercosur (UN, 2019). These percentages may be partly biased due to the unequal number of member countries in these various trade agreements. Indeed, one could make the hypothesis that the share of internal trade within an agreement increases with the number of member countries. However, this argument does not stand when considering empirical facts: NAFTA only has three members and its internal international trade is much higher than that of ASEAN, which has 10 member countries. The reason for these differences in the percentage of internal trade is undoubtedly to be found in the varying degrees of complementarity between member countries’ economic structures.

22Reversely, a high volume of international trade (as a percentage of the countries’ total trade) can exist between countries that have not necessarily signed an agreement. Many so-called “regional” agreements, most of them commercial (over 400 agreements notified to the WTO in 2016), are purely formal and have no effect on the geography of trade. Numerous empirical case studies confirm the discrepancy between regionalization and regionalism. We know for example that for years, the most part (and often the majority) of the foreign trade reported by the EU’s neighboring countries has been with EU countries. The EU’s influence on trade flows predates the signing of any trade agreements under the European Neighborhood Policy. Similarly, ASEAN member countries trade more with East Asian countries (China, Japan, Korea) than within ASEAN.

An improved definition of macro-regional integration

Macro-regional integration according to geographers

23In geography, the region is generally defined on a sub-state level. However, the idea of a regional system can be transposed to groupings of all sizes, which means, for example, that there can be global regions or macro-regions (Pinchemel, 1997; Didelon, 2013). In a geographical sense, macro-regional integration refers to the creation, development and enhancement of interactions between initially distinct but contiguous territorial units located in various neighboring countries. This process ultimately evolves towards the construction of spatially continuous multi-state autonomous regional systems. Interactions between constituent spatial units within these large units are stronger than those with spatial units located outside. The macro-regional entities thus constructed are positioned between the state and the global scale. They can associate several entire contiguous countries, or only portions of these countries’ territories (cf. figure 2).

24In geography, a macro-region is not a purely formal construct, necessarily and solely constructed in an institutional manner – i.e. based on a regional agreement between the governments of neighboring countries. A macro-region is above all a multi-state social construct, produced by the interwoven practices of actors of all kinds and all levels: those can include governments but also businesses, civil society organizations, individuals, networks, entrepreneurial diasporas, students, etc. It can therefore be useful to observe macro-regional integration at the interface of top-down and bottom-up processes, steered from above by governments and driven more spontaneously by other actors as part of horizontal trans-state relations. Taking the analysis a step further, we can define macro-regional integration in a more specific manner:

25• First of all, this expression can either refer to a process (integration in progress) or to a state of affairs (integration accomplished).

26• Second, integration can be sectoral or global: in the first case, it applies to one or more target sectors (for example trade in goods, mobility of people, transport networks, etc.); in the second, it affects a growing number of areas until it encompasses the entire social field in the broadest sense.

27• Thirdly, the macro-regional system that results from the integration process forms a continuous spatial whole that encompasses entire territories or parts of the territories of neighboring countries.

28Integration is complete when there are no longer any barrier effects within the macro-regional system. In theory, the system is characterized by its great internal spatial fluidity.

29Typologies of macro-regional systems can be established using geographical criteria: polarized monocentric systems, polycentric systems, etc. For instance, we could emit the hypothesis that the macro-regional system formed by the EU and its neighboring countries is both polycentric and monocentric: on the one hand, the EU is a complex construct without any real dominant center; on the other hand, because of its size, the EU clearly has a functionally polarizing effect on its immediate environment.

30Finally, there are non-systematic links between regionalism, regionalization and macro-regional integration. The regional concentration of trade can indeed be the result of regionalism (a form of regional multilateralism initiated by neighboring countries that decide to intensify their trade and cooperation). However, it can also be a visible sign of the construction of an integrated macro-regional system.

There is no example of a completed macro-regional integration process in the world

31Two authors who are not geographers but have a keen interest in space have addressed the theme of new regionalism (not just macro) in a geographical and dynamic sense. B. Hettne and F. Söderbaum (1998) present an analysis based on the notion of the region as a social construct. According to these authors, integration consists of moving up one degree on the scale of “regionness”. Integration is the process by which a portion of space, regardless of its size, is gradually “filled” with enough economic, institutional, political, cultural or identity substance to become a distinct system and eventually be perceived as such by its own inhabitants as well as by others. In 1998, B. Hettne and F. Söderbaum identified three degrees of regionness: the “pre-regional” stage where no regions are visible yet but rather what the authors call potential regions or proto-regions; then the stage where macro-regional integration begins in diverse forms (inter-governmental regional cooperation, deeper government-led integration, functional integration driven by the market or caused by the rise of certain social practices in diverse fields); finally, the formation of a sufficiently established region that can become an “acting subject” with its own identity, capacity for political action, and institutional structures whose decisions are regarded as legitimate in the eyes of a transnational civil society. In other words, integration is complete when a regional unit becomes a society. In 2000, the same two authors identify five degrees in the scale of regionness: the region as a geographical unit, the region as a social system, the region as a basic security mechanism, the regional organization and finally the regional society produced by this organization. However, the fundamental idea is the same. The great merit of these authors is that they demonstrate that it is possible and desirable to study integration in a multi-scalar and comprehensive way – that is, by taking all sectors into consideration (Hettne, 1999) – and by giving practices a prominent place. A region is not something that is already there and simply needs to be discovered. It is a dynamic construct, whose existence is not inevitable because integration is a reversible process.

32In the light of this analysis, two hypotheses can be developed. First of all, no macro-regional integration process (in the sense of a multi-state functional system) has ever been fully completed anywhere in the world. The most advanced experiment in this respect is the European Union. However, while it forms a large single market, it is not a perfectly integrated entity. The ultimate degree of regional integration has not been reached: national forces remain powerful within the perimeter of the Community; internal border effects persist; the member countries’ combined populations do not form a society; European identity does not supersede national identity; the EU is not an actor in international relations in many areas and is hardly perceived as such by European citizens themselves or by those of third countries. Secondly, in light of the geographical distribution of exchanges of people and goods, it can be hypothesized that the E.U. does not form a macro-regional system on its own. In this part of the world, it is likely that the space formed by the European Community and many of its neighboring countries is encompassed within a broader macro-regional system (Beckouche, Richard, 2013). Such a hypothesis also appears to be true in Southeast Asia. In ASEAN, trade is scarcely regionalized, since the internal goods trade only accounts for 23% of the total international trade of member countries. In contrast, the region has intense trade exchanges with China, Japan, South Korea and Taiwan. We also know that China’s southern provinces (Guangxi and Yunnan) maintain close functional links with the Indochinese Peninsula (Fau, Franck, 2019).

Macro-regional integration: a methodological gap

33Capturing, measuring, evaluating and comparing macro-regional integration processes is a daunting problem for researchers, posing difficulties that cannot always be resolved. Macro-regional integration is a complex object which, in its most accomplished form, takes into account criteria such as the emergence of a regional identity or the sharing of common social representations, and which may only integrate portions of national territories: when faced with such an object, data and traditional dichotomies (qualitative / quantitative approaches) are not the researcher’s best allies. The analysis of macro-regional integration certainly requires considerable resources, including harmonized protocols to obtain comparable results across all the territorial units observed. It is also certain that given the current state of research on macro-regional integration, we are currently proceeding by trial and error and have difficulty in establishing a universal method.

34From a quantitative perspective, most of the databases that measure the regionalization of trade, investment, migration, student flows, etc., are available at the state level. However, some topics are poorly or not at all documented, and many countries have weak or non-existent statistical services, leaving large portions of the world map blank. These gaps make it challenging to build up global databases of flows between countries; in some cases, they can also make it impossible to construct origin/destination matrices of flows and interactions that would at least provide an accurate cartographic representation of regionalization. For example, a global matrix of a seemingly well-documented theme such as international migration would present many shortcomings, which would make it almost impossible to use. In the field of trade, some data is available beyond traditional databases, in customs registers for example, but it is rarely communicated and not always usable as such.

35In a context where functional macro-regional integration sometimes only affects part of the territory of one or more neighboring countries, the issue with sources becomes even more acute: the databases that provide information on the international movement of people, goods, capital, services, etc., rarely exist at the level of sub-state territorial units. In almost all cases, it is impossible to know whom a given region (or city) within a country trades with. At a first glance, one might think that this difficulty posed by data collection scales can be overcome by working on a geographically restricted monographic case study, to construct a more qualitative regional matrix. However, this option comes with a risk of misinterpretation because in this case, no comparison can be made with other sets. The benefits of building an isolated regional database remain limited, precisely because the integration or regionalization of an ensemble can only be identified if one is able to look at the broader picture to access the general context of reference, move from one scale to another and establish comparisons with other parts of the world.

36Similar challenges arise when adopting a qualitative approach based on field surveys, direct counts (within a restricted perimeter), the observation and analysis of actors’ practices, etc. One then comes up against the contradiction posed by the a priori selection of a space where integration or regionalization is presupposed: once again, this postulate can only be proven if this space is placed in a broader context, which can only be done by using quantitative methods. Nevertheless, multi-site field surveys remain a necessity to observe the empirical reality of processes, question the actors on their feelings and choose the most relevant places of observation, such as certain international borders which remain laboratories of cooperation between territories and societies, and provide points of connection that help us understand neighborhood links and multiple territorial affiliations.

37Given the current state of available data, it must be recognized that integration or regionalization cannot be accurately measured at all levels, in all areas and in all parts of the world. This may lead researchers to instead focus on “indicators of regionality” borrowed from both qualitative and quantitative methods (De Lombaerde, 2006; Deblock, 2017). New regionalism, which derives from a constructivist approach to the region, depicts integration as a rise in degrees of regionness or regionality. From this perspective, the region, whatever its size, is a social construct and each degree of regionality reflects an increasing density of social interactions within this space. This can range from simple intergovernmental regional cooperation to informal integration through the market, to the formation of a region that becomes an actor in international relations, that has its own identity and is perceived as such. In theory, the method for measuring integration is therefore simple: one only needs to define and calculate indicators to reflect each degree of regionality.

38However, the production of a benchmark indicator, such as those provided by the World Bank or UNCTAD on a sectoral basis, is often subject to criticism. Obstacles to be overcome include the choice of variables, the distribution between quantitative and qualitative data, the method of field data collection, the level of cooperation in the states that are required to report the data, etc. Nevertheless, the usefulness of such indicators for the purposes of comparison is proven. In the study of macro-regional integration in particular, comparison and measurement go hand in hand. Their combination is of great methodological importance because it is necessarily through this method that this process, in constant evolution over time and space, can be observed and understood. A common measurement indicator could make it possible to take both of these aspects into account. This is a project that must be opened up and become part of the agenda of researchers who analyze the regionalization of the world.

Conclusion: of the importance of understanding the world’s regional dimensions

39We are going through a period of globalization: an era of all-encompassing connection, networks, global metropolises and the archipelago economy. Economic circuits bypass state networks, whose borders have long since been surpassed by trade flows. However, relations between political, social and economic actors remain partly determined by geographical proximity. Certain spatial constraints, including distance, continue to shape exchanges. Some authors even express skepticism as to the very idea of globalization. B. Hettne and F. Söderbaum (2000) believe, for example, that the social sciences must go beyond what they call the “mystifications of the concept of globalization”, which should not be taken at face value. Indeed, there is much evidence to suggest to the careful observer that global governance is withering away. In this context, an intermediate scale is emerging between the local/national and the world: that of the multi-state macro-region.

40The regionalization of the world and macro-regional integration offer a frame of interpretation that is more in tune than globalization with some of the contemporary world’s political, economic and social realities. This approach is all the more important that this scale may well be the most relevant for dealing with some major issues: regional cooperation, whatever its degree, may for instance prove effective in areas such as international migration management, environment, security, energy, planning, transport, etc., for two reasons. First of all, the issues and challenges associated with these fields of action do not stop at the borders of states because they are transnational in nature, which makes it necessary to take advantage of geographical proximity. Secondly, when a deeper degree of cooperation is envisaged, it can be easier to implement between neighboring countries whose inhabitants share certain preferences and behaviors. The proximity of values, sometimes correlated with the intensity of social and cultural relations and interactions, can facilitate the implementation of common standards, which is a condition for deeper integration.

41Because it places space and the region regarded as social constructs at the heart of its analysis (Lefèbvre, 1974; Paasi, 2010), geography can make a powerful contribution in clarifying key notions such as macro-regional integration, regionalism and regionalization, whose definitions remain fuzzy in a significant number of scientific works. Nevertheless, to find its place in this field, geography must attempt to break down silos by developing interdisciplinary research initiatives. Geographers could contribute their knowledge of regions, their ability to work at different scales, their intimate relationship with space and fieldwork, their methods (quantitative and qualitative) and their know-how in cartographic representation. There are many opportunities for meetings and exchanges with economists, politicians, historians, international lawyers and sociologists. The reflection engaged in a number of recent works aims to stabilize the vocabulary in use, and to develop methods and tools to capture and measure macro-regional integration. This article forms part of this effort to clarify notions and engage geographers in the field of macro-regional studies, also referred to as area studies, where the analysis of networks, circulations of all types and contact spaces is at the core of a relational approach to territories (Brenner, 2004; Lewis and Wigen, 1999). Beyond scientific positioning, macro-regional analysis and action provide an opportunity to refresh a number of economic, political and environmental practices, in a world where not everything is globalized.